What Peroxisomes (Don’t) do to Mitochondria

A well-established function of peroxisomes is the degradation of very-long-chain fatty acids (VLCFA, fatty acids with a carbon chain of more than 20), which do not serve as substrates for the mitochondrial carnitine shuttle (CPT1) (Houten and Wanders, 2010). Instead, VLCFA are imported into peroxisomes via ABCD transporters, where they undergo chain shortening by peroxisomal β-oxidation enzymes, such as acyl-CoA oxidase 1 (ACOX1). The resulting shortened acyl-CoAs can be shuttled by carnitine-O-octanoyltransferase (CROT) and carnitine-O-acetyltransferase (CRAT) to mitochondria for further oxidation to acetyl-CoA (Violante et al., 2013). When peroxisomal β-oxidation is impaired – as in PBDs – VLCFA accumulate in cells and tissues. Diagnostic evaluation of these disorders often includes measuring the plasma ratio of C24:0 and C26:0 to C22:0 (Rattay et al., 2020). While VLCFA accumulation has well-documented effects on cellular lipid homeostasis – including altered membrane composition and lysosomal function (Kleinecke et al., 2017) – growing evidence suggests that mitochondria are also affected.

One of the diseases most commonly associated with VLCFA buildup is X-linked adrenoleukodystrophy (X-ALD), a genetic disorder caused by mutations in the ABCD1 gene, leading to defective VLCFA transport into peroxisomes. The contribution of VLCFA to the disease phenotype in X-ALD has been under debate, specifically whether VLCFA accumulation directly causes neurodegeneration or acts through secondary pathways (Kemp and Wanders, 2010; Martinović et al., 2023). Intriguingly, more recent findings link VLCFA to altered mitochondrial dynamics. In fibroblasts from X-ALD patients, VLCFA accumulation leads to increased mitochondrial ROS production and oxidative stress (López-Erauskin et al., 2013). A recent study by Launay et al. (2024) demonstrated that hexacosanoic acid (C26:0) induces the phosphorylation of dynamin-related protein 1 (DRP1), a master regulator of mitochondrial fission that is also termed DLP1 or DNML1. This phosphorylation promotes excessive mitochondrial fragmentation and morphological abnormalities, further implicating mitochondrial dysfunction in the pathology of peroxisomal disorders. The mechanism likely involves depolarization of the mitochondrial inner membrane, disrupted ATP synthesis, and activation of apoptotic pathways.

The relevance of these findings is supported by earlier studies in rat hippocampal glia and neurons, where VLCFA exposure was shown to induce mitochondria-mediated cell death involving membrane depolarization and oxidative stress (Hein et al., 2008). These studies underscore that VLCFA toxicity is multifaceted, with mitochondria acting as key mediators of the cellular stress response. While much clinical focus has historically centered on reducing VLCFA levels – such as in the case of Lorenzo's oil, a dietary therapy developed by the parents of an X-ALD patient – the mitochondrial consequences of VLCFA accumulation remain a vital area of research, particularly for developing effective early interventions.

Beyond VLCFA, recent work has revealed that medium- and long-chain fatty acids (MCFA, LCFA) are also relevant in the context of peroxisomal dysfunction. In Drosophila melanogaster, RNAi-mediated knockdown of dABCD1 results in neurodegeneration in photoreceptor cells, which can be rescued by dietary supplementation with MCFA (Gordon et al., 2018). Similarly, Pex3 mutants, which lack a peroxisome biogenesis factor, show a lipid profile skewed toward longer-chain triglycerides at the expense of MCFAs (Faust et al., 2014). Another study in mice revealed that the peroxisomal enzyme L-PBE degrades dicarboxylic acids derived from dietary MCFAs, an essential process for liver health (Ding et al., 2013). We quantified fatty acid methyl esters in Drosophila mutants that lack the peroxisome assembly factor Pex19, which results in early lethality. We found that VLCFA were present only in the pmol/mg range, whereas M- and LCFA were present in nmol/mg concentrations (Sellin et al., 2018). These mutants accumulated VLCFA and showed a marked depletion of M- and LCFA, shifting the average fatty acid chain length toward longer species. This metabolic imbalance coincided with hyperactivation of mitochondrial β-oxidation, mediated by Hepatocyte Nuclear Factor 4 (HNF4) (Bülow et al., 2018; Palanker et al., 2009). Ultimately, we showed that this hyperactive mitochondrial β-oxidation results in MCFA depletion, and that dietary MCFA supplementation reduces mitochondrial ROS production and rescues the lethality of Pex19 mutants. This highlights a previously underappreciated regulatory interaction between peroxisomes and mitochondria, in which peroxisomes not only process lipids but also modulate mitochondrial metabolic activity and stress responses.

Peroxisomes are involved in α-oxidation, the degradation of branched-chain fatty acids (BCFA), notably phytanic acid, which is derived from the chlorophyll-borne diterpene alcohol phytol found in the diet. Phytol is obtained primarily through dairy products, ruminant animal fats, and vegetables. Bell pepper is a particularly rich source of phytanic acid as phytyl esters (Krauß et al., 2016). Phytol is converted to phytanic acid through a series of reactions. The α-oxidation of phytanic acid requires hydroxylation of C_α by phytanoyl-CoA hydroxylase (PHYH), followed by decarboxylation, resulting in pristanic acid, which subsequently undergoes β-oxidation (Wanders et al., 2023). Defects in α-oxidation, often due to mutations in PHYH, lead to the accumulation of phytanic acid, which can be toxic to cells. Toxicity particularly affects mitochondria, where it disrupts membrane integrity and function. The accumulation of phytanic acid and its derivatives can thus lead to impaired mitochondrial β-oxidation, reduced ATP production, and increased oxidative stress due to the generation of ROS (Busanello et al., 2012; Grings et al., 2012; Zemniaçak et al., 2023). These mitochondrial dysfunctions become manifest in various clinical symptoms as promotors of disease, but not as the sole cause. Particularly affected are tissues with high energy demands such as the nervous system and muscles, contributing to the clinical presentation of Refsum disease, a peroxisomal Hereditary motor and sensory neuropathy (HMSN), caused by variants in PHYH or the peroxisome import receptor PEX7.

Peroxisomes are involved in the biosynthesis of plasmalogens, glycerophospholipids that were first described a century ago in 1924. Plasmalogens contain a vinyl ether group at the sn-1 position, the first carbon of the glycerol. Plasmalogens are rich in polyunsaturated fatty acids (PUFAs) and represent a surprisingly large proportion, roughly one fifth, of membrane phospholipids in the heart and brain (Braverman and Moser, 2012). Substrate fatty alcohols are generated from acyl-CoA by peroxisomal fatty acyl reductases, and two further steps in plasmalogen biosynthesis are catalyzed by peroxisomal glycerone-phosphate O-acyltransferase (GNPAT) and alkyl-dihydroxyacetone phosphate synthase (AGPS), respectively. Biosynthesis is completed in the endoplasmic reticulum (ER). Plasmalogens are oxidation-sensitive due to the vinyl-ether bond and the PUFA content, and may thus act as ROS-scavengers (Sindelar et al., 1999). The involvement of plasmalogens in the execution cell death by ferroptosis appears dependent on cell-type, context, degree of unsaturation and other factors (Zheng and Conrad, 2024). Plasmalogens may thus grant protection from oxidation, but may also act as supporters of ferroptosis.

Plasmalogens are present in the mitochondrial membranes, making up between 10 and 40% of their phospholipids and supporting the high curvature of the cristae of the inner mitochondrial membrane (Venkatraman et al., 2024). Defects in the mitochondrial phospholipid transacylase tafazzin cause a severe form of cardiomyopathy termed Barth syndrome (BTHS) with striking alterations in cardiolipin structure as well as choline plasmalogens in the heart and ethanolamine plasmalogens in the brain (Kimura et al., 2019). The importance of peroxisomal plasmalogen metabolism in the heart and in other mitochondria-rich organs is underscored by a recent meta-analysis of mRNA and protein expression in various tissues (Plessner et al., 2024). The study finds that plasmalogen biosynthesis, and, more generally, PTS2/PEX7-dependent pathways are more highly expressed in tissues rich in mitochondria such as the heart.

An adipose-specific mouse knockout of Pex16 provides evidence that peroxisomes contribute to mitochondrial integrity through plasmalogens. In the knockout, temperature-dependent fission that normally occurs in the cold, is inhibited, and mitochondrial function is consequently compromised. Dietary supplementation of alkylglycerols as plasmalogen precursors partially restores mitochondrial morphology and function (Park et al., 2019), supporting the notion of a role of peroxisome-borne plasmalogens on mitochondrial health.

Taken together, plasmalogens appear to belong to the factors that mitochondria require from peroxisomes so that in cases of peroxisome deficiency, mitochondrial damage can result from plasmalogen loss. To what extent this functional interaction of peroxisomes and mitochondria depends on plasmalogens, is currently not known.

Electron transfer is a key energy-metabolic process in every cell and is regulated by reduction-oxidation (redox) reactions (Lismont et al., 2015). Loss of redox equilibrium leads to either reductive or oxidative stress, both of which are detrimental and associated with a broad array of diseases. During redox reactions, reactive oxygen and reactive nitrogen species (ROS/RNS) are formed. At physiological levels, they have important signaling functions; for example, hydrogen peroxide (H₂O₂) acts as a proliferation signal and controls axon pathfinding in retinal ganglion cells of zebrafish (Sies, 2017). We showed that mitochondrial H₂O₂ drives the activity of dopaminergic neurons in Drosophila (Paradis et al., 2022). Both mitochondria and peroxisomes are important sources of ROS. In mitochondria, ROS such as superoxide anions form at the electron transport chain during respiration. Likewise, in peroxisomes ROS generation is part of their normal function: during their metabolism, H₂O₂ is generated by oxidases and scavenged by catalase and peroxidase. A recent study proved that peroxisomes can take up mitochondrial ROS at ACBD5/PTPIP51-mediated contact sites to alleviate mitochondrial stress (DiGiovanni et al., 2025). Peroxisome metabolism significantly contributes to overall cellular H₂O₂ levels. For instance, induction of peroxisomal β-oxidation through fibrates (PPARα agonists) or other peroxisome proliferators increases H₂O₂ production, often exceeding the detoxification capacity of catalase and leading to oxidative stress (Ivashchenko et al., 2011; Reddy and Hashimoto, 2001; Schrader and Fahimi, 2006). Furthermore, peroxisomal dysfunction due to catalase inhibition or genetic defects in peroxisome biogenesis leads to intracellular accumulation of H₂O₂ and downstream oxidative damage (Lismont et al., 2019).

While both organelles have antioxidant systems to balance ROS formation, their enzyme complements are distinct. Mitochondria primarily contain superoxide dismutase 2 (SOD2), whereas catalase is localized almost exclusively to peroxisomes. SOD1 is found in the cytosol but also enters peroxisomes via copper chaperone for SOD1 (CCS), despite lacking a canonical peroxisomal targeting signal (Islinger et al., 2009). In mammalian cells, SOD2 is exclusively mitochondrial (Karnati et al., 2013), while catalase is delivered to peroxisomes through a non-canonical peroxisomal targeting signal (PTS1), which has lower affinity for the Pex5 import receptor (Legakis et al., 2002).

During aging, catalase import into peroxisomes becomes less efficient, resulting in H₂O₂ accumulation and contributing to cellular senescence. Although catalase is not naturally mitochondrial, overexpression of mitochondria-targeted catalase has shown a protective role in specific disease models. For example, mitochondrial catalase protects against radiation-induced damage and reduces oxidative stress in aging muscle, improving outcomes in models of sarcopenia and neurodegeneration (Liao et al., 2013; Parihar et al., 2015; Xu et al., 2021). These findings underscore the organelle-specific detoxification capacity and suggest that shifting antioxidant capacity toward mitochondria can be therapeutically beneficial in oxidative stress-related conditions.

Peroxisome dysfunction has profound effects on mitochondrial redox balance. Inhibition of catalase or impairment of peroxisome biogenesis induces mitochondrial ROS generation, including superoxide anions (^·O₂^-), thereby contributing to mitochondrial oxidative damage and bioenergetic decline (Bülow et al., 2018; Lismont et al., 2019). This crosstalk between the two organelles is not only spatial but functional, with peroxisomes buffering cytosolic H₂O₂ levels and modulating redox-sensitive mitochondrial pathways. The redox interaction involves complex signaling and metabolite exchange, including the transfer of fatty acid intermediates and NAD⁺/NADH redox equivalents (Fransen et al., 2013; Schrader et al., 2015). These studies highlight that the integrity of peroxisomal function is essential for maintaining mitochondrial redox homeostasis and overall cellular oxidative balance.

Like mitochondria, peroxisomes contain dehydrogenases and oxidases that use the oxidized form NAD⁺ of the redox pair NAD⁺/NADH for oxidation of substrates. In mitochondria, reduced redox equivalents resulting from mitochondrial oxidation processes can be reoxidized directly in the mitochondrial respiratory chain. This option does not exist in peroxisomes, as no regeneration system appears to be present within peroxisomes. Also, there is no transporter for the direct shuttling of NADH/NAD⁺. Until recently, it was mechanistically unclear how a redox equivalent shuttle between peroxisome and cytosol or mitochondria might work.

Through a combination of experimental and bioinformatic approaches, it was found that both a subunit of lactate dehydrogenase and cytosolic malate dehydrogenase encode functional peroxisomal targeting signals in a region that was originally assigned as 3'-UTR (Hofhuis et al., 2016; Schueren et al., 2014; Stiebler et al., 2014). During translation, the stop codons of these two gene products are interpreted as sense codons (functional translational readthrough) with a frequency of 2–8%, which leads to an extension of the original subunits of LDH and MDH (Schueren and Thoms, 2016). The protein extensions encode peroxisomal targeting signals, so that part of the normally cytosolic LDH and MDH is imported into peroxisomes, explaining how the two abundant dehydrogenases can get access to the peroxisome.

We hypothesized that concomitant presence of LDH (and MDH) in the peroxisome and in the cytosol could form the basis for redox shuttle systems mediating redox transfer between these compartments. This shuttle would use the peroxisomal LDH for the formation of lactate from NADH and pyruvate, thereby regenerate NAD⁺. Lactate is exported to the cytosol, and further transported to mitochondria. Reimport of pyruvate, generated by cytosolic LDH, closes the cycle. The LDHB readthrough is conserved in all mammals, while the MDH1 readthrough is conserved in all vertebrates. Interestingly, the PTS1 in non-mammalian vertebrate MDH1 is more efficient than the mammalian MDH1 (Hofhuis et al., 2016). This led to the prediction that the functions of peroxisomal LDH and MDH can complement each other (Hofhuis et al., 2016), a hypothesis that was recently confirmed: Redox shuttling is only disrupted if both readthrough extensions are deactivated. However, this mechanism does not seem to be required for peroxisomal β-oxidation (Chornyi et al., 2023), so that the exact biological functions of the peroxisomal readthrough dehydrogenases remain to be elucidated. The underappreciated readthrough and redox connection between peroxisomes and mitochondria adds another layer of metabolic connection between these organelles.

Membrane contact sites (MCSs) between mitochondria and peroxisomes are specialized zones where the two organelles are closely apposed, enabling the exchange of metabolites, coordination of lipid metabolism, and integration of redox signaling pathways essential for cellular homeostasis.

In a screen for organelle contact sites in yeast, Shai et al. (2018) identified two tethering mechanisms between peroxisomes and mitochondria: one involving the peroxisomal membrane protein Pex34 and an (as yet) unidentified mitochondrial partner, and another involving the mitochondrial mitofusin Fzo1, possibly through homotypic interaction. A recent study showed that Fzo1 localizes to peroxisomes, where it can interact with mitochondrial Fzo1 (Alsayyah et al., 2024). Whether a similar interaction of Mitofusin 1 (Mfn1) on both peroxisomes and mitochondria constitutes a functional tether between the two compartments is still elusive, but proximity labeling with peroxisomal membrane proteins revealed the proximity of both Mfn1 and 2 in mammalian cells, suggesting that this ER-mitochondria tether has a similar function in peroxisome-mitochondria contact sites (Huo et al., 2022).

The protein ACBD2 (also known as ECI2) has been identified as a tether that mediates physical interactions between peroxisomes and mitochondria in Leydig cells, facilitating the formation of MCSs and supporting inter-organelle communication (Fan et al., 2016).

Peroxisomes and mitochondria share many functions (fatty acid oxidation, ROS homeostasis) and proteins. Under peroxisome dysfunction, some membrane proteins such as PEX13 or PEX14 mislocalize to mitochondria. The path of such protein mislocalization may be related to the mechanism of peroxisome biogenesis, especially the targeting of peroxisome membrane proteins. Peroxisomes have been suggested to be hybrid organelles of ER and mitochondrial origin (Sugiura et al., 2017). Some membrane proteins are transported through the ER (Hoepfner et al., 2005; Kim et al., 2006; Thoms et al., 2012). Peroxisome- and mitochondria-derived vesicles (MDV) have been reported to fuse to become mature, import-competent organelles (König et al., 2021; Sugiura et al., 2017). However, many factors in the targeting of peroxisomal and mitochondrial membrane proteins, such as similarity in targeting sequences, dual or sequential organelle targeting by N-terminal mitochondrial targeting signals and C-terminal peroxisomal targeting signals, protein misfolding and aging can contribute to mistargeting of peroxisomal proteins to mitochondria. Mutations in peroxisome assembly factors can thus lead to a retention of other peroxisomal proteins in mitochondria.

Recently, the E3 ubiquitin ligase MARCH5 (also termed MITOL) has been shown to play a role in mitochondria-dependent peroxisome biogenesis (Verhoeven et al., 2024; Zheng et al., 2025). The enzyme has previously been implicated in pexophagy and mitochondrial quality control including mitophagy (Zheng et al., 2022). Upon mitochondrial depolarization, MARCH5 translocates to peroxisomes, a process dependent on Parkin-mediated ubiquitylation and the activity of peroxisomal biogenesis factors PEX3 and PEX16. Using single and double knockouts of MARCH5 together with peroxisome biogenesis factors, two studies have provided evidence for a role of MARCH5 in peroxisome de-novo formation by Pex3-dependent budding from mitochondria (Verhoeven et al., 2024; Zheng et al., 2025). The redistribution of MARCH5 from mitochondria to peroxisomes, along with its role in peroxisome formation, supports the hybrid model of peroxisome biogenesis, which proposes membrane protein trafficking from both, mitochondria and the ER, to peroxisomes.

We proposed that mislocalization of peroxisomal proteins to mitochondria contributes to mitochondrial pathology in peroxisomal disorders (Thoms et al., 2009). Indeed, in PEX3-deficient cells, overexpression of the mitochondrial AAA-type ATPase ATAD1 ‒ known for its ability of removing mislocalized proteins from the membrane ‒ partially recues the mitochondrial phenotype (Nuebel et al., 2021). Currently, it is not clear if this effect is due to re-localization or removal of peroxisomal (membrane) proteins. Moreover, we need to understand if such protein mislocalization is generally present in PBD, because PEX3, together with PEX19 and PEX16 are not only the rarest causes of PBD (Thoms and Gärtner, 2012), these three proteins are also the only peroxins known that lead to an apparent absence of peroxisomal membranes.

Peroxisomes were shown to localize to Mitochondria-ER Contact sites (MERCs), both in yeast and mammalian cells (Cohen et al., 2014; Guillén-Samander et al., 2021). Contact sites between mitochondria and the ER are among the best studied MCSs. Their major functions are phospholipid and calcium ion exchange, and about 20 proteins with tethering function have been described (Bülow and Sellin, 2023). MERCs are implicated in various neurodegenerative diseases, prominently Parkinson's disease.

The lipid transport protein VPS13D tethers both mitochondria and peroxisomes to the ER by bridging of Miro, a protein of the outer mitochondrial membrane and VAP proteins at the ER (Guillén-Samander et al., 2021). Miro has a splice variant that causes its Pex19-dependent enrichment at the peroxisomal membrane, and expression of this splice variant leads to recruitment of VPS13D to peroxisomes. This might provide the cell with a mechanism to transfer lipids from the ER to both peroxisomes and mitochondria by their respective variants of Miro, raising the idea of a triple or three-way connection between peroxisomes, mitochondria, and the ER.

Evidence for such triple contacts with a role in lipid homeostasis comes from a study describing wrappER, rough ER in mouse hepatocytes that wraps around peroxisomes while establishing contact sites with mitochondria (Ilacqua et al., 2022). Fatty acid β-oxidation occurs at high rates in hepatic triple contacts in both peroxisomes and mitochondria, but the contacts between mitochondria and wrappER are regulated during fasting-to-feeding transition while wrappER-peroxisome contacts remain stable. At 3 h postprandial, peroxisomes are released from the triple contact. These triple organelle contacts probably provide hepatocytes with a fine-tunable mechanism to regulate fatty acid flux.

Mitochondria and peroxisomes interact functionally and defective peroxisomes contribute to mitochondrial dysfunction (Figure 1). In normal physiology, plasmalogen biosynthesis requires peroxisomes, as does the degradation of VLCFA and BCFA. When peroxisomal function is impaired, the absence or accumulation of these lipids can lead to mitochondrial damage. Peroxisomes also rely on redox equivalent shuttling with mitochondria, likely involving lactate-pyruvate and malate-aspartate shuttle systems. Defects in peroxisome function trigger the accumulation of mitochondrial ROS and oxidative damage, which might be due to upregulated mitochondrial FAO. Some PMPs are reported to reach the peroxisome through MDVs. In peroxisome biogenesis deficiency, some PMPs accumulate in mitochondria. To address the pathologies found in PBD and peroxisome-associated disorders, it will be important to dissect the individual contributions of individual metabolic and protein factors in the interaction of peroxisomes and mitochondria.

We thank Nina Bonekamp, Markus Islinger, and Britta Nemeita for comments on the manuscript.